Target Tracking Technology And Its Application And Research In Ultrasonic Detection Image

In modern industry,because of the defects of industrial equipment parts,the operation failure often occurs or even causes more serious consequences.Therefore,the nondestructive testing of industrial equipment parts is particularly important.At present,nondestructive testing in our country mainly depends on many years of testing experience of testing personnel to judge the defects of parts,which leads to the subjectivity of testing and the accuracy of testing can not be guaranteed.In this paper,based on the full study of the target tracking technology,the multi-target tracking algorithm is applied to the defect tracking of ultrasonic detection image,which is of great practical significance to improve the automation degree,detection efficiency,reduce the subjective components in the process of ultrasonic detection and improve the detection accuracy.In this paper,the target tracking algorithm is studied and improved.Finally,the most suitable target tracking algorithm is selected to be applied in nondestructive testing.The specific work is as follows:(1)Aiming at the shortcoming that the space-time context tracking algorithm cannot automatically detect the location and size of the tracking target,this paper studies an automatic detection and tracking algorithm based on inter-frame difference and space-time context based on the space-time context tracking algorithm.The obtained foreground target information and the outline of the target are transmitted to the STC tracker in order to track the target of the next frame.Comparative experimental results show that the improved algorithm can realize the automatic detection and tracking of moving targets,and improves the tracking accuracy.(2)In view of the shortcomings of the compressed perceptual tracking algorithm,which is easy to lose the tracking target when the target moves too fast and the tracking accuracy is not high,this paper makes two improvements under the framework of the compressed perceptual tracking algorithm: Firstly,the weighted function is introduced to assist in judging the positive and negative samples of the target,and the unreliable samples are eliminated in time,so as to solve the problem of the low accuracy of the compressed perceptual tracking algorithm in complex background environment.Secondly,when the moving target moves too fast,the kalman filter algorithm is used to predict the position of the current frame of the target,so as to solve the problem that the target is lost due to the target moving too fast.The experimental results show that the improved algorithm can track the moving target stably in the complex background environment,and will not lose the target.Finally,the improved algorithm is used to track the ultrasonic detection image,which proves that the target tracking technology can be effectively applied to the ultrasonic detection image.(3)In the actual nondestructive testing work,there is more than one defect in the same section of the parts,and the single target tracking algorithm is difficult to meet the actual work requirements.Therefore,this paper focuses on the kalman multi-target tracking algorithm based on the background target,and applies the algorithm to the tracking of the ultrasonic testing image.Through a large number of experiments,it is proved that the algorithm can be realized Automatic detection and tracking of multiple defect targets in ultrasonic image provide a reliable guarantee for the subsequent 3D reconstruction of the defect center of parts,and lay a foundation for the automation of nondestructive testing.(4)In this paper,the above-mentioned multi-target tracking algorithm is applied to the ultrasonic nondestructive testing experimental system established by the project team in the early stage to automatically detect and track the defect target in the ultrasonic image.The experimental system can automatically detect the defects of parts,then automatically detect and track the defects in the collected ultrasonic image,and finally reconstruct the center position of the defects.The experimental results show that the average point error of 3D reconstruction is 1.0307 mm,the maximum average error of single defect target is 1.3186 mm,and the average angle deviation of fitting line is less than 4.6158°.The effectiveness of multi-target tracking algorithm is verified by the analysis of point error and line error of 3D reconstruction,and the application of multi-target tracking algorithm improves the automation of nondestructive testing.